水稻遥感估产信息系统拥有包括地理坐标数据、各种光谱数据在内的各种数据,作者提出了对二维地理空间进行编码,由一个编码代替两个地理坐标,从而实现对地理坐标数据的压缩,其压缩率为1:2。文中还针对土壤光谱特点,利用实验室标准土壤光谱与实际土壤光谱之间的关系,提出了土壤光谱回归压缩方法,将每条土壤光谱曲线,压缩为2个回归系数。 The rice remote sensing yielding information system has various data including geographic coordinate data and various spectral data. The author proposes to encode the two-dimensional geographic space, and to replace the two geographical coordinates by one code so as to compress the geographical coordinate data , The compression rate of 1: 2. In this paper, the spectral characteristics of soil were also analyzed. Based on the relationship between laboratory standard soil spectra and actual soil spectra, a regression method of soil spectral regression was proposed. The spectral curves of each soil were compressed into two regression coefficients.